When it comes to presidential comparisons, scholars always struggle with a small-n problem — there is only one president at a time, so we look across presidencies in the full awareness that the historical circumstances may change quickly. I was quite explicit about my periods, the past 32 years versus the 20 years before that. I did not compare the recent trend to the 1800-1824 period other than to note that the last time presidents had enjoyed such success was in the early years of the republic, and I pointed out the very different partisan circumstances then and now. (An aside: no, Truman did NOT win reelection. He completed FDR’s fourth term and won on his own in 1948.)

I’m not sure why a Republican winner in 2016 would be more likely to face a primary challenge. Neither Reagan nor Bush 43 did. I don’t propose an explanation for primary challenges (or the lack thereof) to sitting presidents. Given the small-n problem already discussed, that might be a theory with too few cases to test all the variables.

Your n is too small. So small that it makes me wonder which explanation has more validity: the one that you provide here, or the simple alternative that there is in fact no clear pattern.

(2) What constitutes your bounded period of comparison?

You need to clarify and justify the “beginning” of your period under consideration. Is it Clinton 1992? Kennedy 1960? FDR 1932 (who succeeded himself twice, which you don’t mention)? For example, if you include all president’s going back to the immediate postwar period, the picture looks suddenly different because Truman and Eisenhower both won reelection,and the Kennedy case is neutral (or at least unknowable) because of the assassination. In fact, if you were comparing periods, you could have noted that FDR, Truman, and Eisenhower each won reelection, and that this would likely be the most appropriate point of comparison.

(3)Given the rise of the modern presidency, can you justifiably compare across time in the way that you do when you position Thomas Jefferson, James Madison, and James Monroe up against Clinton, GWBush and Obama?

Especially so given the novel modern advent of the presidential primary system, which means incomparable modes of nominating to Jefferson et al. Also, what about the rise of the unilateral powers of the presidency through control of modern administrative capacity? The ability of modern presidents to reward friends and punish enemies is much greater in the modern period. But this is less significant as a factor than lack of primary challengers why?

(4) Isn’t the lack of viable third party candidates (a la Ross Perot) just as credible a variable as the lack of party primary challenger to a sitting president?

That is, if we are listing the absence of a factor as a serious (and causal) variable.

(5) Can you really boil down such a complicated multicausal phenomenon to a single factor (lack of primary challenger-> advantageous early nomination) that is internal to the political system alone?

Essentially, you have held the entire world outside of the American political system “constant” from Jefferson to Obama.

(6) Final deep thought: Isn’t this just globalization run amok?

Certainly a fascinating question though.

“If this pattern holds in the future, the stakes rise in the next presidential election. With no incumbent on the ballot, we will not be choosing a president for four years, but very likely for eight.”

Not necessarily. The tendency toward party unity is not equal between the two parties. A Republican winning in 2016 is much more likely to face a rebellion from the Right in 2020 than a Democrat would be to face a rebellion from the Left.

Moreover, a third-party intrusion is more likely to be right-populist than left-populist.

Broadly speaking I think you’re right. If the electorate is potentially more dynamic, it should generally be to the advantage of the candidate (incumbent or challenger) more likely to lose under static conditions. Maybe it is just a coincidence that the last two presidential incumbents happened to have some public opinion disadvantages that they were able to overcome (partially) as a result of this. Also the Democratic Party did happen to overtake the Republicans organizationally between 2004 and 2008 (Kreiss 2012 makes this point well, I can’t believe I just gave a cite in the comments section of a blog post; I’m a dork).

But I also think it’s worth considering if the recent success of incumbents is in anyway due to an intrinsic advantage that they have in an environment more conducive to mobilization; and I could at least at speculate on why this may be the case: (1) I think the bully pulpit-however weakened it may be- is still probably the best mechanism for broadcasting the uptake of new issues to non-voters (who usually do not follow politics closely)-for example Bush and gay marriage in ’04 as a way to appeal to evangelicals, and Obama and Hispanics with immigration, also following (2) you could probably also use executive orders to give deliverables to appeal to these groups in ways that a challenger couldn’t (i.e. Obama and the DREAM Act stuff).

As far as (1): I don’t think thisdiscourages primary challenges, but is a complimentary theory that may also help explain why incumbents have seemed to do better recently. I would agree if anything it would seem to encourage primary challenges, but there is really nothing to test this against yet, as both Bush and Obama (despite inevitable discontent from their sides ideologues) were pretty popular with their parties rank and file.

I can see how these new technologies and strategies helped Bush and Obama get re-elected, via boosting base turnout. But there are 2 related things re: Andy’s argument that don’t make sense to me here:

1) How does this discourage, as opposed to encourage, a primary challenge? If, as you say, these technologies lower both costs and efficiency (of both communication, and generating turnout), then it could potentially be easier, not harder, to mount a nomination challenge — i.e. to create buzz, get the candidate on ballots, and then work (especially caucus) turnout, etc. I’m imagining a larger-scale example of what happened to former Senators Bennett and Lugar.

2) In a general election, again, how are these tools necessarily an advantage for the incumbent? Both sides can mine all this data, then tweet, toot, and Facebook away, to max out their target voters. And in fact, both campaigns attempted to do so in 2012. Either the Obama base was just naturally larger, or his campaign was just better at the game, or both.

My point is, Axelrod and Plouffe, et al, were great at this when Obama was a candidate, and were still great when he was President. This is not an advantage of incumbency.

No I would more or less agree with this, although I would also suggest another important factor is how technological shifts have impacted electoral strategies in regards to mobilization; and maybe give incumbents more paths to victory in difficult environments.

If you look at the Bush and Obama reelection campaigns, both candidates spent most of the campaign season underwater in their approval ratings, yet both campaigns counteracted this by focusing less on winning over the median ‘undecided’ voter than mobilizing non/sporadic voters among their respective bases. Bush did this through micro-targeting evangelicals, while Obama used digital technologies to help build a huge volunteer-based field apparatus.

Newer technologies have both helped give campaigns more information about the public as a whole, as well as lowering the costs of communication within it. As a result, parties are able to more efficiently mobilize new or sporadic voters. This may allow an incumbent facing strong headwinds against them to be less dependent on traditional swing voters, and give them the flexibility to pursue a turnout strategy among segments of the electorate that might have been previously to difficult to mobilize in the numbers necessary.

I think this explains a good deal of how Bush and Obama won,and maybe means incumbents in this era will have a little more agency in campaigning in difficult electoral environments, but I guess we will see about the generalizability moving forward…

Yet still, I wonder about a few things here:

1) Wasn’t having to face a primary challenger (as Ford, Carter, and HW Bush did) more of a symptom, than a cause, in those 3 cases? In other words, sure, there is no shortage of journalistic accounts pinning their losses (to one degree or another) on bruising primary challenges (Carter in 1980 especially). But it seems to me that all 3 of them were near certain general election losers anyway.

2) Might Citizens United potentially throw a wrench into this pattern, at some point? For instance, let’s imagine the Republicans nominate and then elect a more moderate candidate in 2016 (a Mitch Daniels, say), who governs from the center-right. All it would take is a handful of multimillionaires, to mount even a halfway credible primary challenge on the order of Buchanan 1992. (replace Gingrich/Adelson w/ serious candidate, and less shady billionaire).

Just some thoughts. Ultimately, the pattern you suggest sounds correct. But we would need a circumstance in which things are really going horribly wrong for a sitting president, be able to test it. I mean, think of the circumstances (across the board, in domestic and foreign policy), when Kennedy challenged Carter in 1980.

Jonathan Keller